Full text of "Herd Organization and Movements of Elk in Wind Cave National Park, South Dakota"

HERD ORGANIZATION AND MOVEMENTS

OF ELK
IN WIND CAVE NATIONAL PARK, SOUTH

DAKOTA

FEDERAL
PUBiJCAT'ON

L lY L

Natural Resources Report
Number 13

U.S. Department of the Interior
National Park Service

Digitized by the Internet Archive

in 2013 with funding from

LYRASIS IVIembers and Sloan Foundation

http://archive.org/details/herdorganizationOOherd

Herd Organization and Movements of Elk
In Wind Cave National Park, South Dakota

Kenneth L. Varland

Graduate Assistant

Iowa Cooperative Wildlife Research Unit

Iowa State University, Ames

Allan L. Lovaas

Wildlife Biologist

Wind Cave National Park

Hot Springs, South Dakota

Robert B. Dahlgren

Leader

Iowa Cooperative Wildlife Research Unit

Iowa State University, Ames

U.S. DEPARTMENT OF THE INTERIOR • NATIONAL PARK SERVICE
NATURAL RESOURCES REPORT NUMBER 13 • 1978

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402

ABSTRACT

Herd organization, movements, and distribution of eli< at Wind Cave National Parte were
studied from June 1973 to February 1975, to aid in management of this nonmigratory
population. Thirty-two elk were marked with either ear flags, colored collars, or radio
collars. Six calves were captured by hand; the remaining 26elk were marked by immobilizing
them with succinylcholine chloride shot from a dart gun (20 adult bulls in a baited corral
trap, 4 adult cows and a 6-month-old female calf from a helicopter, and I free-ranging adult
cow on the ground).

Marked elk were observed 713 times and located 1 13 times by telemetry. The fenced.
I 14 km- (44-square mile) park accommodated three relatively discrete cow-calf herds, with
each herd using a distinct area of the park. The largest, about 170 elk, occupied the
northwestern 20.7 km- (8 square miles) (Beaver Creek region) and did not intermingle with
about 90 elk occupying 25.9 km- (10 square miles) to the east (Boland Ridge region) although
their ranges overlapped slightly. About 40 elk occupying the southwestern 1 1.7 km- (4.5
square miles) (Gobbler Knob region) occasionally intermingled with the largest herd for
brief periods in January and February 1974. and a few crossed the west fence to spend
spring, summer, and early fall of 1974 in a 18. 1 km'-^ (7-square mile) area of the Black Hills
National Forest. Within each herd, cow-calf groupings each changed in individuid com-
position with time. Movements of marked bulls were variable, but most remained in the
northwestern portion of the park close to the trap site. Individual bulls interchanged freely
between small groups of bulls from one day to the next. From limited evidence, it seems
likely that the bull population is also divided into three discrete herds.

Cows and calves used the following areas most intensively: west of Red Valley in the
wooded region adjacent to Highland Creek in the southeastern portion of the park: on and
east of Boland Ridge; the areas in, adjacent to, and between Sanctuary and Research
Reserve prairie dog towns; and areas adjacent to and including Cold Brook Canyon in the
southwestern corner of the park. Bulls were seen most often in the northern half of the
park.

Elk were observed most readily during the hours closest to sunrise and sunset. They
generally fed in grassland areas and bedded in wooded areas. Elk usually avoided use of
steep slopes in all seasons. East- and south-facing slopes were used more than west- or
north-facing slopes during most of the year.

Each herd should be managed individually to prevent an overuse of the range in any one
area of the park. Information from this study was used in elk trapping operations during
January 1976 and 1977. Elk numbers in the Gobbler Knob region have been declining in
recent years because of hunting outside the park and it was unnecessary to trap elk from
this region. Reduction quotas were set for the Boland Ridge and Beaver Creek cow-calf
herds; no more than the stated quota could be taken from each herd. Although 319 elk
were removed from the population in 1976 and 1977. the quota for either herd was never
exceeded.

If needed in the future, suggested locations for an additional trap to capture cows and
calves from the Beaver Creek and Gobbler Knob regions were discussed. Adult bulls from
the Beaver Creek region can be captured by baiting them into the present trap, but additional
traps would be needed to capture bulls from the Boland Ridge or Gobbler Knob regions.

Ax the Nation's principal conservation a,i>enc\\ the Department of the Interior has re-
sponsibility for most of oar nationally owned public lands and natural resources. This
includes f>sterin)> the wisest use of our land and water resources, protecting our fish and
wildlife, preserving the environmental and cultural values of our national parks and his-
torical places, and providing for the enjoyment of life through outdoor recreation. The
Department assesses our energy and mineral resources and works to assure that their
development is in the best interests of all our people. The Department also has a major
responsibility Jbr American Indian reservation communities and for people who live in
Island Territories under U.S. administration.

INTRODUCTION

The American elk or wapiti (Cervus elaphus) was
the most widely distributed member of the deer family
when Europeans first came to this continent (Murie
1951). As settlers moved westward, this animal dis-
appeared from most of its original range. By 1900, only
small relict herds remained, largely in the Rocky
Mountains and along the Pacific Coast.

Elk were native to the area now known as Wind
Cave National Park (WCNP) but were virtually elim-
inated by market hunting before 1900 (Hipschman
1959). The park, established in 1903, is a sanctuary for
many species of wildlife native to the Great Plains.
Lovaas (1973a) summarized the history of elk herd
reestablishment and management at WCNP, beginning
in 1914 when 21 Rocky Mountain elk (C. e. nelsoni)
were transplanted from Jackson Hole, Wyoming, to
a fenced 1 ,683-ha (4,160-acre) game preserve within
the park. A transplant of 25 additional elk from Yel-
lowstone National Park in 1916 supplemented the pop-
ulation. After a time, the ever-increasing herd was re-
duced, primarily through shooting by park rangers, and
the meat donated to Indian people. Attempts to bait
or drive elk out of the park through openings in the
boundary fence were ineffective in controlling popu-
lation growth, and some of the elk driven out may have
returned to the park.

The park was enlarged to its present 11,355 ha
(28,059 acres) in 1946. During the winters of 1953-54
and 1954-55, 1,000 elk, from an estimated population
of 1,100, were shot. Herd reduction by shooting was
terminated after 1957 because of adverse public re-
action to large-scale shooting of elk in Yellowstone
National Park.

The wildlife management plan for WCNP, adopted
in 1964, recommended that the population be main-
tained at 150-300. In the fall of 1969, the elk population
was estimated at 800. Consequently, from 1 969 to 1972,
657 elk were trapped and transplanted into neighboring
Custer State Park and the Jicarilla Apache and Ogalala
Sioux Indian reservations.

Using helicopters, the population was reduced by
driving elk into the park's corral trap for bison (Bison
bison), which was modified to accommodate elk. In-
formation from field reconnaissance, censusing, and
trapping indicated that elk concentrated in three main
areas within WCNP. Susceptibility of elk to capture
differed greatly among areas of the park; a helicopter
could not be used successfully to drive elk over the
rough terrain, heavily forested area, and long distances
to a single trap near the north boundary. To maintain
numbers compatible with the carrying capacity of the
range, refinements in elk trapping were necessary for
long-term management of this nonmigratory popula-
tion. The purpose of this study, therefore, was to pro-
vide information about herd organization, distribution,
and movements of WCNP elk to effect overall park-
wide reductions without the excessive reduction or
elimination of some population segments and insuffi-
cient reductions of others. Problems in this study were
similar to those faced by Craighead et al. (1972) in
Yellowstone National Park and Smuts (1974) in South
Africa.

Intensive field work at the park was conducted by
Varland from June 1973 through August 1974. Lovaas
made periodic field trips through February 1975, and
a few observations were made by various park em-
ployees from February to October 1975.

STUDY AREA

Wind Cave National Park is located in the south-
eastern foothills of South Dakota's Black Hills. The
park is bounded on the east and south by private lands,
on the west by the Black Hills National Forest, and
on the north by Custer State Park (Fig. 1). WCNP lies

in the transition zone between the confierous forests
and grasslands of the state (Shult 1972).

The northwestern part of WCNP is largely
north-south wooded ridges that are hilly to steep, ex-
cept for gentle slopes in the upland valleys. The re-

Figure I. M;ip of Wind Cave National Park.

National Park Service

mainder of the park is in the prairie areas of the Black
Hills footslopes, and major physiographic features are:
(1) Red Valley, a trough-like area in the eastern portion
of the park; (2) the Sandstone Hogback, a steep, an-
gular escarpment east of Red Valley with a wooded
top that is rolling to hilly and is notched by the heads
of steep-walled canyons that drain to the east; and (3)
the Limestone Plateau, located between Red Valley
and the ridged area, which is gently sloping to hilly
except for southeast flowing drainages and their trib-
utaries, which have cut deep, wooded valleys and can-
yons (U.S. Soil Conservation Service 1969). Eleva-
tions range from 1,111 m (3,646 ft) msl in Red Valley
to 1,528 m (5,013 ft) in the northwestern corner.

Mixed grass prairie covers about three-fourths of
WCNP (Fig. 1), and the predominant grasses include
western whealgrass {A g ropy ran smithii), needlegrasses
{Stipa spp.), common buffalograss (Buchloe dacty-
loidcs), bluestems (Andropogon spp.), bluegrasses
(Poa spp.), grama grasses {Boutelouu spp.), and Jap-
anese brome {Bromiis japonicus). Common forbs of
the park include fringed and cudweed sagewort (Ar-
temisia frigida and A. ludoviciana), slimflower and
silverleaf scurfpea (Psoralea tenuiflora and P. argo-
phylla), upright prairieconeflower (Ratibida colum-
naris), purple prairieclover {Petalostenwn purpur-
eum), American vetch (Vicia americana), dotted
gayfeather (Liatris punctata), sedges {Carex spp.),
blacksamson echinacea {Echinacea angitstifolia), hairy
go\denaster {Clirysopsis villosa), and heath aster {Aster
ericoides). The major browse species of WCNP is true
mountain mahogany {Cercocarpus montanus) but other
shrubs found in the park include skunkbush sumac
{Rhus trilohata), rose {Rosa spp.), leadplant amorpha
(Amorpha canescens), western snowberry {Symphor-
icarpos occidentah's), currant {Ribes spp.), common
chokecherry (Prunus virginiana), and American plum
{P. aniericana).

The remainder of the park is covered by ponderosa
pine (Pinus ponderosa) found primarily in the north-
western portion. Much of the pine cover is in a sa-
vannah-like configuration, whereas some north-facing
slopes have highly concentrated stands of young trees.

A detailed description of predominant range sites
and vegetation was given in a conservation plan pre-
pared for the park by the U.S. Soil Conservation Ser-
vice (1969) and summarized by Lovaas and Bromley
(1972). Plant names are from Beetle (1970), except that
we preferred cudweed sagewort over Louisiana
sagewort.

Average annual precipitation is about 46 cm (18
inches), nearly 70% of which falls between 1 May and
30 September. Snow cover is usually light and
intermittent.

The park is surrounded by a 244-cm (8-ft) woven
wire fence, except for a 4.8-km (3-mile) segment of
122-cm (4-ft) fence along the western boundary that
permits egress of elk to national forest lands. The only
interior fences are those that surround the park's head-
quarters and campground; thus the elk can move freely
throughout the park. Two creeks flow into WCNP,
Highland Creek from the north and Beaver Creek from
the west; both go underground within the park bound-
aries (Fig. 1). Spring-fed concrete dish tanks and res-
ervoirs are located at various places in the park to
provide additional water for wildlife. The road network
includes two highways, four improved dirt roads, and
26 active jeep and (or) hiking trails (Fig. 2).

Blacktail prairie dog {Cynomys ludovicianus) towns
(Fig. 3) occupied nearly 445 ha (1,100 acres) of prairie
area in 1971 (Lovaas 1973b; Petersburg 1973). Lovaas
( 1973b) reported that these prairie dog towns have been
expanding rapidly since 1955. WCNP is a sanctuary
for several other species of wildlife, including the fol-
lowing herbivores: bison, pronghorn antelope {Antil-
ocapra aniericana). mule deer {Odocoileus liemionus),
and white-tailed deer (O. virginianus).

MATERIALS AND METHODS

The principal technique used to determine elk herd
organization and movements in Wind Cave National
Park was that of observing movements of marked in-
dividuals; however, all observations of elk were re-
corded to determine the most heavily used ranges of
WCNP.

The emphasis of the study was on the capture and

marking of elk from cow-calf groups because those
animals constituted the largest proportion of elk cap-
tured and removed in the park's helicopter-trapping
program and because they were also sources of new
calf crops. However, information concerning adult bull
movements is also valuable, and attempts were made
to mark those elk as well.

Natural Resources Report No. !3

Highway

Improved dirt road

Jeep and (or) hiking trail

Figure 2. Network of roads and trails in Wind Cave National Park.

North Boun dary ( 16 acresl
O

Pringle Cul oil
(?2 acres)

Rankin Ridge
(11 acres)

Sanctuary
( 176 acres)

Highland Creel<
( 10 acres )

Cold Brook Cany(
(32 acres)

Boland Ridge
(4 acresl

Fimiri- 3. Prairie dog towns in Wind Cave National Park in 1971 (Lovaas 1973b; Petersburg 1973).

National Park Service

Elk Observations

Initially, the park was divided into three general re-
gions (Fig. 4) which in the past had winter elk con-
centrations. These were the Gobbler Knob region, in-
cluding Wind Cave Canyon and all areas south of the
forested region just west and northwest of the park
headquarters; the Beaver Creek region, including most
of the area north of Beaver Creek and west of Highland
Creek in the northern half of the park plus the area
about I mile south and paralleling Beaver Creek; and
the Boland Ridge region, including all areas east of
Highland Creek and the adjacent forested areas west
of Highland Creek in the central and southeastern por-
tion of the park.

Generally, observations of elk were made from a
motor vehicle or on foot during early morning and later
afternoon hours by using either a 15-60 x spotting
scope or 7 X 35 binoculars for visual relocations of
marked elk; a radio receiver was used to locate radio-
equipped animals. Radiotelemetry locations were made
by triangulation as described by Heezen and Tester
(1967).

With the aid of a U.S. Geological Survey map of
WCNP (scale 1:24,000), locations of elk were recorded

to the nearest 16.2 ha (40 acres) by using a six-digit
code. The first two digits represented the range and
township, the third and fourth, the section number,
and the final two, the coordinates for the location of
elk within a section. A clear-acetate-grid overlay di-
viding a section into 16 numbered portions of 16.2 ha
(40 acres) each was used to record elk observations on
the map.

The park's two highways, four improved dirt roads,
and 26 active jeep and (or) hiking trails were traveled
in attempts to cover the regions where elk generally
were concentrated. Four hundred and five different
observation routes were taken, including off-the-road
hikes. The areas of the park observed from each route
were delineated and divided into 16.2-ha (40-acre)
tracts. No systematic pattern of routes was used; how-
ever, all portions of the park were visited regularly.
Russo (1964) pointed out advantages of not using a
systematic pattern of established routes for counting
deer in Arizona. Dusek ( 1 975) also used this technique
of regular but nonsystematic visits for studying com-
petition between deer and cattle on ranges in Montana.

Information recorded for each field trip included
observation route taken as well as the following data

Gobbler Knob
Region

GcbDIer

Knob

Figure 4. The three general regions of Wind Cave National Park.

Natural Resources Report No. 13

for each elk group observed: time, location, activity,
group size, sex, age, vegetation type, degree of slope,
and slope exposure.

Groups of elk were classified as either bull groups
or cow-calf groups. Bull groups consisted of adult
males only. Cow-calf groups included adult females,
yearling females, calves, and yearling males. Adult
bulls seen with cows and calves were excluded in de-
termining sizes of cow-calf groups. Yearling males
were included in the cow-calf category because 79%
of all yearling males observed were with cows and
calves.

Seasons corresponded with general climatic condi-
tions: spring — March, April, and May; summer — June,
July, and August; fall — September, October, and No-
vember; and winter — December, January, and Feb-
ruary. Data for the same seasons of different years
were combined.

Vegetation types were classified as forest or grass-
lands. Degree of slope was estimated as level to gentle
(0-25°) or medium to steep (>25°). Slope exposures
were classified according to the four cardinal points
(N, S, E, W) or as level ground (having little or no
slope).

To determine which general areas of WCNP were
used most heavily by elk, we divided the park into
quarter-sections (64.7 ha or 160 acres). We summed
the numbers of marked and unmarked elk seen in each
quarter-section, and then divided that sum by the total
number of trips to each 16.2-ha (40-acre) tract within
each quarter-section to obtain an average number of
elk per trip to each quarter-section. The number of
observation trips to each tract was recorded (an ob-
servation trip means only that a tract was visible from
a route and does not mean that we necessarily traveled
through it). The average numbers of elk per 100 trips for
the quarter-sections were arrayed and mapped to show
areas of primary and secondary use. If the number of
elk seen in a quarter-section was equal to or above the
midpoint of the array, that quarter-section was desig-
nated primary range; if the number was below the
midpoint, that quarter-section was designated secon-
dary range. Separate designations were made for each
season and each type of elk group. Elk were more
difficult to observe in forested than in other areas and
thus were perhaps underrepresented. Hikes through
forested regions were made regularly to help alleviate
this problem.

Capture of Elk

Searches for newborn calves were made in early
morning and late afternoon almost daily during the
calving seasons of 1973 and 1974. The calving season
in WCNP is believed to extend from the last week in

May through the third week in June. Johnson (1951)
reported a calving period of 21 May to 12 June, with
the peak on 1 June, for the Gallatin herd in south-
western Montana. Late calving was suspected in Mon-
tana when two calves, which were much smaller than
others, were observed in mid-July. Schwartz and
Mitchell (1945) reported a calving period of 15 May to
15 June, with a peak from I to 10 June, for Roosevelt
elk in Washington. They recorded a birth as late as 10
July.

Calves were secretive during most of the summer
and were not seen as often as older elk. No so-called
calving grounds (Brazda 1953) were found in the park,
although more calves were seen in certain areas than
in others (Fig. 5).

An average of one calf per 4.2 hours was seen during
the estimated 138 man-hours spent in ground searches
for calves from 9 to 29 June 1973. One male and two
females were captured (one female escaped before
being marked) for an average of one capture per 46
man-hours.

From 27 May to 28 June 1 974, an estimated 279 man-
hours was spent in ground searches, with an average
of one calf observed per 3.9 man-hours. Ground
searches in 1974 resulted in the capture of two females
and one calf where the sex was not determined. An
average of 93 man-hours was spent for the capture of
each calf.

Most calves chased on foot escaped easily because
they could outrun pursuers. Two calves were captured
as a result of chases during ground searches; both went
into the prone position after being chased a short dis-
tance. The remaining calves stayed in the prone po-
sition from the time they were first sighted until the
time of capture. Bedded calves generally became
frightened and ran away when a searcher inadvertently
walked close by. One of those that remained down was
captured.

In 1974, early morning searches for calves were made
with a helicopter. Nursery groups were located from
the air, and attempts were made to split individual
calves away from the group and then force the fright-
ened calf into the prone position. When this occurred,
the passenger disembarked, and the helicopter hovered
over the calf until the capture was made.

Forty calves were seen during 5 hours spent in hel-
icopter searches from 19 to 21 June 1974; an average
of 8 per hour. Only one male was captured. More
calves may have been captured if the helicopter had
been used earlier in the month uhen calves were
younger.

Older elk were captured by immobilizing them with
succinylcholine chloride shot from a dart gun on the
ground, from a helicopter, or in a corral trap.

National Park Service

^6 973 Do

o *6 12 73 □

6-19 74*

Gc bDler
A Knob

oCalf sighting

* Date and location
of coif capture

Figure 5. Locations of elk calves sighted or captured from 9 to 29 June 1973 and from 27 May to 28 June 1974
at Wind Cave National Park.

The least successful ground-capture method used
was the driving of elk by hunters on foot; only an adult
cow and a yearling bull were hit by a dart with this
method (Varland 1976b). Stalking was the most suc-
cessful, with hits on five adult cows and one adult bull.
Two adult and two yearling bulls were hit from natural
blinds (large rocks, trees, etc.). Elk that escaped into
rough, wooded areas after they were hit were difficult
to follow or find. Only one cow was marked success-
fully by using a ground immobilization method (stalk-
ing). Although ground methods were relatively inef-
fective, a large number of observations was made of
both marked and unmarked elk during those activities,
thus the time was not wasted.

Adult bulls often entered the corral trap, which was
baited with salt; 23 were captured from October 1973
to January 1974, and 2 in the summer of 1974. This
does not include a bull that escaped before the gate
was closed and two that jumped the corral fence after
being trapped.

No cows and calves were attracted to the baited
trap, but cow-calf groups seldom were seen within
1.6 km of the structure. In previous years, cows and
calves were captured when they wandered into the trap
(Lovaas 1973a). thus cows and calves may have been

captured during this study if traps had been located
within the areas they used.

Not counting original construction cost, the corral
method was the least costly technique used to capture
and mark elk ($24.75 per elk). Use of a helicopter was
more effective and less expensive ($280.20 per elk) than
blind, drives, or stalking ($2,454.00 per elk) for im-
mobilization of free-ranging elk (Varland 1976b). Costs
include salaries, vehicle and helicopter expenses where
applicable, and costs for immobilization equipment
(does not include the cost of drugs and marking
devices).

Marking of Elk

Thirty-two elk were marked during the study: 20
bulls, 5 cows, 6 newborn calves, and a 6-month old
female calf. Belt collars were placed on 9 bulls, and
rope collars on 10 bulls and I cow. One bull, 4 cows,
and the 6-month- old calf were equipped with radio
collars. Only ear flags were attached to newborn
calves.

Marked elk were assigned an individual code number
to denote where and when an elk was marked
(BR=Boland Ridge, BC^Beaver Creek, GK =
Gobbler Knob, and CT= corral trap). For example.

Natural Resources Report No. 13

elk BR3 was the third elk captured and marked in the
Boland Ridge region.

Colored ear flags (3.75x46 cm or 1.5x18 inches),
used to mark individual calves, were constructed from
Saflag material (Safety Flag Company of America,
Pawtucket, R.I.). Metal ear tags were inserted through
slits in the folded Saflag, and the entire marking device
was then attached to the calfs ear with ear-tag pliers.
One flag was attached to each ear.

Older elk were marked with collars ranging from 78.7
to 83.8 cm (31 to 33 inches) in circumference (Knight
1966), depending on sex and age. Adult bulls were
fitted with either a belt collar, a rope collar, or a radio
collar. Cows were equipped with either a radio collar
or a rope collar.

Colored belt collars were made by sewing Saflag
material onto 3-ply rubberized belting 5 cm (2 inches)
wide (Electrical Engineering and Equipment Com-
pany, Des Moines, Iowa), in designs similar to those
described by Progulske (1957). The ends of the collar
were joined together with size 25E alligator steel-belt
lacing (Electrical Engineering and Equipment Com-
pany) into which a 10-penny nail was placed and bent
over.

Rope-type collars similar to those used by Craighead
et al. (1969) were made by inserting six pieces of col-
ored Saflag (7.6x22.9 cm or 3x9 inches) through half-
inch polypropylene rope at 10.2-cm (4-inch) intervals
in color combinations that enabled us to recognize in-
dividuals at sight from either side. The two ends of the
rope collar were joined by using four No. 3 hog rings.

Six radio-transmitter collars (150-151 MHz) were
purchased from Wildlife Materials, Inc., Carbondale,
III. Visual recognition of three radio-equipped elk was
made possible by either sewing Saflag strips (3.75 cm
or 1.5 inches wide) to the collars, or by wrapping col-
ored tape (Scotch brand colored cellophane. No. 650,
Minnesota Mining and Manufacturing Co., St. Paul,
Minn.) around them at 10.2-cm (4-inch) intervals.
Three collars were marked with Saflag and three with
colored tape. The ends of each collar were joined by
using belt lacing and a nail.

Visibility of collars was good throughout the study.
Saflag was more durable and more visible than the
colored tape on radio collars. Bulls with collars were
not individually recognized in 18% of the sightings be-

cause of poor light, long distances, or rising heat waves.
Rope collars were more readily recognized than belt
collars. The maximum distance at which a collar was
individually recognized through a 15-60 x spotting
scope was approximately 3.2 km (2 miles) for a rope
collar and 2 km ( 1 .25 miles) for a belt collar. Optimum
conditions were necessary for recognition of collars at
distances greater than 0.8 km (0.5 mile). Collars gen-
erally were recognizable under relatively poor condi-
tions at distances of 0.4 km (0.25 mile) or less. Two
belt collars (white with yellow stripes and white with
yellow dots) were sometimes difficult to differentiate.

One belt collar (blue with green dots), recovered
from an adult bull found dead (probably poached) about
17.5 months after it was marked, was worn but still
recognizable at 0.8 km (0.5 mile), through a 20 x spot-
ting scope.

The longest periods that collars were known to be
recognizable on a living elk were 451 days for a rope
collar and 438 days for a belt collar. However, we do
not know how long the collars would retain a recog-
nizable pattern.

Ear flags could be seen with the naked eye at dis-
tances up to 0.8 km (0.5 mile). Ear flags on certain
individuals were recognizable up to 1.6 km (1.0 mile)
when using a spotting scope.

Ear flags were fairly durable, but some were even-
tually lost. The two calves marked in 1973 retained
both flags until late November 1974. when each lost
one. One of the calves marked in 1974 was never re-
located, one lost an ear flag after about 1 month, one
after about 5 months, and one retained both up to the
end of field work. One ear tag and the attached flag
(elk BC4) was sloughed about 1 month after the calf
was marked and was found later in the summer in good
condition, with ear cartilage still inside the tag.

The known transmitting life for radio collars ranged
from 27 to 351 days, with a mean life of 186 days. This
does not include one radioed cow that was never re-
located either visually or by radio telemetry after it
was marked; this cow's radio had been activated 39
days before it was attached to the cow and was working
properly at the time of attachment. One radio trans-
mitted properly for the first 83 days after activation,
then stopped for at least 1 84 days, and then transmitted
again for 46 days before it stopped permanently (per-
haps because of battery failure).

National Park Service

HERD CHARACTERISTICS

Co w -Ca If Herds

Areas used by cows and calves were determined by
mapping all observations of individuals marked in a
particular region of the park and then connecting the
outermost points (Fig. 6). The cow-calf population was
divided into three relatively discrete herds. Elk of the
Boland Ridge region occupied an area of 25.9 km^ (10
square miles) in the eastern portion of WCNP. Beaver
Creek elk occupied about 20.7 km^ (8 square miles) in
the park's northwestern corner. Gobbler Knob elk
spent most of their time during winter in a 11.7-km'^
(4.5-square-mile) area in the park's southwestern cor-
ner. The area in the park occupied by the two elk
marked in this region varied from 23.3 to 28.5 km- (9
to 1 1 square miles), depending on whether one includes
brief winter movements to the Beaver Creek region
and other areas in the Gobbler Knob region. Gobbler
Knob elk occupied an additional area of about 18 km'-
(7 square miles) in the Black Hills National Forest west
of the park from early April to early October 1974.

Cow-calf groups were largest during winter months,
and group size was highly variable (Table 1). During

ground counts in the winter of 1973-74, the largest
group seen, 140, was in the Beaver Creek region, and
groups ranging from 51 to 120 were observed there
frequently. The largest group seen in the Boland Ridge
region was 80 elk, and groups ranging from 40 to 70
were seen commonly. Groups of from 20 to 46 were
seen frequently in the Gobbler Knob region.

The same trend in group size was observed for the
winter of 1 974-75, with the largest groups in the Beaver
Creek region and smaller groups in the Boland Ridge
and Gobbler Knob regions. The largest groups ob-
served in the Beaver Creek, Boland Ridge, and Gob-
bler Knob regions were 162, 91, and 41, respectively.
Elk marked in a particular region usually were seen
associated with the larger groups of that region. Calf
production probably accounted for the increased group
sizes from one winter to the next within a particular
region, except in the Gobbler Knob region where elk
numbers decreased. The probable reason for the de-
cline of elk in the Gobbler Knob region, hunting outside
the park, will be discussed later. The mean ratio of
calves per 100 cows in the park during July, August,

Figure 6. Areas used by all cows and calves marked in each region in Wind Cave National Park.

Natural Resources Report No. 13

Table I. Cow-calf groups and bull groups seen during the four
seasons at Wind Cave National Park.

Total

No. of

Season

elk

groups M

ean size ± SD

Range

Cow-calf groups

Spring

2,937

195

15.1 ±

20.7

1-99

Summer

3.423

311

11.0±

17.3

1-127

Fall

4.874

213

22.9 ±

31.3

1-119

Winter

6,802

162

42.0 ±

48.5

1-162

Year-long

18,036

881
Bull groups

20.5 ±

29.4

1-162

Spring

1,272

279

4.6 ±

4.5

1-17

Summer

2,146

523

4.1 ±

5.1

1-35

Fall

1,493

482

3.1 ±

3.4

1-30

Winter

1,543

331

4.7 ±

5.0

1-33

Year-long

6,454

1,615

4.0 ±

4.5

1-35

Overall

24,490

2.496

9.8 ±

17.8

1-162

and September 1 973 was 34: 1 00. The ratio in the Beaver
Creek region was 29: 1 00 and in the Boland Ridge region
was 49: 1 00. Too few cows and calves were seen in the
Gobbler Knob region during this period in either year
to calculate a meaningful ratio. In 1974, the mean ratio
for WCNP was somewhat higher (43:100) than the
previous year, and ratios in the Beaver Creek region
(40:100) and the Boland Ridge region (46:100) were
comparable.

An aerial elk census was conducted on 13 March
1975. One hundred and twenty-four bulls were ob-
served throughout the park. Only four groups of cows
and calves were seen (Lovaas 1975), including a group
of 36 in the Gobbler Knob region and one of 170 in the
Beaver Creek region. These group counts were similar
to ground counts made in the winter of 1974-75 in those
regions. Two groups of cows and calves were seen
about 1.6 km (1 mile) apart within the Boland Ridge
region, one of 7 and one of at least 28. There could
have been more in the second group because it was
scattered through a wooded area. Lovaas et al. (1966)
reported that counts of elk were most accurate when
there was snow cover and elk were in the open, away
from wooded areas. Therefore, we believe some of the
elk in the Boland Ridge region were not seen during
the aerial census; this explains the discrepancy be-
tween the ground and aerial counts.

Bull Herds

From limited evidence, it seems likely that the hull
population was also divided into three discrete herds.

The areas occupied by these herds differed slightly
from those occupied by cow-calf herds, particularly
in the area south and east of the corral trap.

Bulls marked in the trap interchanged freely between
bull groups from one day to the next throughout the
study, although they generally remained in the north-
western portion of the park (Beaver Creek region);
only a few moved to other regions at various times.
All sightings of 15 of the 20 marked bulls were within
a 6.4-km (4-mile) radius of the trap, and only two were
seen more than 9.7-km (6 miles) from the site. We
inferred that the Boland Ridge and Gobbler Knob re-
gions each had separate bull herds also because (1)
unmarked bulls were seen consistently in those regions,
several of which were identifiable by antler confor-
mation and condition (broken tines, etc.); (2) marked
bulls were rarely seen in either region throughout the
study; and (3) no marked bulls were ever seen in either
region during the rut. It should be noted, however, that
cows from the Gobbler Knob region spent a large por-
tion of the rutting period outside the park, and some
of them may have mated with bulls that lived outside
the park.

Bulls marked in the trap were relocated a total of
321 times. Dispersal from the trap site was varied.
Spring sightings (116) were concentrated south and
west of the trap (Fig. 7). Long-range movements in-
cluded one bull seen east of Boland Ridge near the
park's eastern boundary and another bull that crossed
the park's boundary fence into the Black Hills National
Forest.

Summer sightings (84) generally were located west
of the trap, with another concentration occurring
southwest of the park's fire tower (Fig. 8). A few scat-
tered sightings were made southeast of the trap, in-
cluding one bull seen on Boland Ridge (the same bull
seen there in the spring. Fig. 7). Only one marked
invididual was seen south of Beaver Creek during this
season.

Fall sightings (46) usually were scattered west and
south of the trap site. Movements during this period
may have been related to the rut, because a large
cow-calf herd was in the Beaver Creek region (elk CT6
was seen with a harem in this region in September
1974). Marked bulls were seldom seen east or southeast
of the trap during fall (Fig. 9). One bull sighted in the
Black Hills National Forest during spring (Fig. 7) was
killed in the fall by a hunter about 3.2 km (2 miles)
north of where it first was seen outside the park. An-
other bull was seen as far away as the park's south-
western corner (Gobbler Knob region) in November
1974.

Winter locations (75) were similar to those during
fall, except that more sightings were made east and

National Park Service

Figure 7. Locations of marked adult bulls during spring 1974 and 1975.

Figure 8. Locations of marked adult bulls during summer 1974.

Natural Resources Report No. 13

Figure 9. Locations of marked adult bulls during fall 1973 and 1974.

Figure 10. Locations of marked adult bulls during v\intcr 1973-74 and 1974-75.

National Park Service

southeast of the trap (Fig. 10). Marked bulls were not
observed to make long-range movements to other re-
gions of the park during winter.

Mean sizes of bull groups varied less by season than
did cow-calf groups, but were highly variable within
seasons (Table 1). Mean group size was highest in
winter, but the largest group (35) was seen during sum-
mer. Few bulls were observed in the Gobbler Knob
region, and groups there never exceeded 10 animals.
Group-size characteristics for bulls of the Beaver
Creek and Boland Ridge regions were similar; small
groups of 2-10 bulls were seen most often, followed
by single bulls, and then by groups of 1 1-20. Occa-
sionally, groups of 31-40 bulls were seen in both
regions.

Movements and Interactions of Marked Individuals

Relocations of marked individuals generally con-
firmed the presence of separate herds of elk. Of the
32 elk captured and marked in the park, only two were
never relocated. Eight hundred and twenty-six relo-
cations were made of 30 elk (321 locations of bulls
marked in the corral trap and 505 of elk marked else-
where), including 713 made visually and 1 13 solely by
radio telemetry. Radio-equipped elk sometimes were
located visually after a telemetry location was made.
The following discussion concerns individual move-
ments of marked cows and calves. Locations of indi-
vidual bulls were given by Varland (1976a).

Boland Ridge Region

Two cows and two calves were marked in the Boland
Ridge region, but one cow (BR4) was never relocated.
Elk BR], BR2, and BR3 stayed in the eastern half of
the park and were associated with the herd there that
numbered 80 elk in 1973 and 91 in 1974. Marked elk
generally were found in two areas: (1) on or east of
Boland Ridge, and (2) west of Red Valley in the wooded
region adjacent to Highland Creek in the southwestern
portion of the Boland Ridge region. Cow-calf groups
were seldom seen in Red Valley, a prairie area with
only small patches of wooded cover. Movements of
the three marked elk were similar to each other. No
major seasonal shifts in movement patterns were noted,
except that few relocations were made west of Red
Valley during fall. Elk marked in this region were never
seen with cows or calves marked in other regions of
the park, but commonly were seen together. All three
marked elk were seen together seven times, and on 19
occasions two of the marked elk were seen together,
exclusive of other marked elk. Elk BRI and BR2 were
seen together nine times, BRI and BR3 were seen
together eight times, and BR2 and BR3 were seen to-
gether twice.

Elk BRI, a spike bull marked as a calf on 12 June
1973, east of Boland Ridge, was relocated 57 times
(Fig. 11). This elk was last seen on 7 February 1975.

Elk BR2 was marked as an adult cow on 3 January
1975, on the west slope of Boland Ridge and was re-
located 36 times (Fig. 12). This cow's radio transmitter
did not work most of the time, and visual observations
were made in most instances. This cow was last located
on 24 January 1975.

Elk BR3. a male calf captured with the aid of a
helicopter on 19 June 1974, east of Boland Ridge, was
relocated 26 times and was last seen on 28 February
1975 (Fig. 13).

Beaver Creek Region

Two cows and four calves were marked in the Beaver
Creek region. Elk BC6. a female calf, was never re-
located after being marked. The remaining five marked
elk stayed in or near and ranged throughout the wooded
northwestern portion of the park. These elk were as-
sociated with a cow-calf herd that numbered 140 elk
in 1973 and 170 in 1974. Movements of the marked
animals were similar to each other. No major seasonal
shifts in movements occurred, although few relocations
during summer and fall were made north of a line ex-
tending from the fire lookout tower to the corral trap.
Elk marked in the Beaver Creek region were associated
in different combinations on 45 occasions during the
study (Table 2). Elk BC4 and BC5 were marked late ~
in the study but frequently were seen associated with
three elk marked earlier. Individuals marked early in
the study commonly were seen together at various
times throughout the study. The five elk were never
seen together at the same time.

Table 2. Frequency of observed associations in various combi-
nations involving two, three, or four of the five elk marked
in the Beaver Creek region.

Associations observed

Frequency occurrence

BCl +

BC2

BCI +

BC3

BC2 +

BC3

BC2 +

BC4

BC2 +

BC5

BCI +

BC2

BC3

BCI +

BC2

BC4

BCI +

BC2

BC5

BCI +

BC3

BC5

BCI +

BC4

BC5

BC2 +

BC3

BC5

BC3 +

BC4

BC5

BCI + BC2 + BC3 + BC4
BCI + BC2 + BC3 + BC5

Natural Resources Report No. 13

GcbDief
A Knob

* CapUire site

* Spr ing relocal ions ( 8 )
o Summer r eloca tions (19)
■ Fa I I relocations (10)

* Winter relocations (20)

Figure 11. Locations of elk BRl from 12 June 1973 to 7 February 1975.

The range of elk marked in this region overlapped
slightly with the range used by elk marked in the Boland
Ridge region. However, marked elk from the Beaver
Creek region were never observed mingling with
marked elk from the Boland Ridge region, nor were
they seen in the Gobbler Knob region.

Elk BCI, a female marked as a calf on 19 June 1973,
near Highway 87, was relocated 53 times (Fig. 14), and
was last seen on 10 December 1974.

Elk BC2, an adult cow, was stalked, immobilized,
and marked on 26 November 1973 and was relocated
128 times (Fig. 15). This cow, whose radio transmitter
stopped working in October 1974, was last seen on 10
January 1975.

Elk BC3, an adult cow immobilized from a helicop-
ter, was marked on 18January 1974, about a mile south-
east of the fire tower and was relocated 46 times (Fig.
16). She was last seen on 7 March 1975.

Elk BC4, a calf of undetermined sex, was marked
on 31 May 1974, on the western edge of Research
Reserve prairie dog town (Fig. 3). This elk was relo-
cated eight times (Fig. 17), and was last observed on
26 September 1974.

Elk BC5, a female calf, marked in Curley Canyon
on 5 June 1974. was relocated 16 times (Fig. 18). This
elk was last seen on 5 October 1975.

Gobbler Knob Region

A 6-month-old female calf (GKl) and an adult cow
(GK2) were immobilized and radio-collared in January
1974, in the Gobbler Knob region. The movements of
both were similar. Both elk spent the first winter pri-
marily in the southwestern corner of WCNP in asso-
ciation with a cow-calf herd of 46 elk. During this
period these elk made brief excursions into the Beaver
Creek region. From early April to early October, they
were usually relocated just west of the park in the Black
Hills National Forest. Elk GKl returned to the park
in early October 1974, and was relocated several times
during the winter of 1974—75, in association with a herd
of 41 elk. Elk GK2 was last relocated in mid-August
outside the park. Both marked elk of this region were
relocated together, by visual observation 15 times and
by radio telemetry 17 times. Elk from the Gobbler
Knob region and the Beaver Creek region were seen
associated during January and February 1974, on one
occasion for each of the following combinations: GKl
+ GK2 + BCI + BC2: GKl + BCI + BC2: and GKl
+ GK2 + BC2. The following combinations were seen
during the same period on two occasions: GK 1 + BC2;
GK2 + BCI + BC2: and GK2 + BCI.

Elk GKl, marked as a 6-month-old female on 4 Jan-
uary 1974, in the extreme southwestern corner of the

National Park Service

Gc bDler
A Knob

♦ Caplure sile

* Spring felocalions (10)
o Summer relocations (9)
■ f il I relocations ( 3 )

• Winter relocations (M )

Figure 12. Locations of elk BR2 from 3 January 1974 to 24 January 1975.

GcbDIer
A Knob

♦ Capture site

* Spring relocations (0)
o Summer relocations (10)

• Tal I relocations ( 7 )

• Winter relocations (9)

Figure 13. Locations of elk BR3 from 19 June 1974 to 28 February 1975.

Natural Resources Report No. 13

Gc bDler
■^ Knob

* Caplure site

* Spring relocations (17)
o Summer reloca I ions (1 1 )
■ Fal I relocdiions (11)

* Winler relocations (14)

Figure 14. Locations of elk BCl from 19 June 1973 to 10 December 1974.

Gc bbier
^ Knob

* ("apt u '(? Site

A Spring r elO(, a I ions U1 )

o S'jinnu'r i oloca I ions (3?)

■ Kii I relocalions ( 8 )

• Winie' -elocitions (4?)

Figure 15. Locations ofclk BC2 from 26 Novcnihcr 1973 to 10 January 1975.

National Park Service

F&fK
Hq<

GqbDIer
Knob

+ Capl u re site

* Spring relocations (16)
o Summer r eloca t ions (10)
■ Fdl I relocations ( 6 )

• Winter relocations (14 )

Figure 16. Locations of elk BC3 from 18 January 1974 to 7 March 1975.

GcbDIer
A Knob

* Capture site

* Spring relocations (0)
o Summer reloialions(6)

• I hM relocdi'ons ( 2 )

• Winter relocations ( '

Figure 17. Locations of elk BC4 from 31 May to 26 September 1974.

Natural Resources Report No. 13

Gc bDler
^ Knob

♦ Capture site

* Spring relocations C 1 )
o Summer relocations ( 5 )
■ Fdl I relocations ( 6 )

• Winter relocations (4)

Figure 18. Locations of elk BC5 from 5 June 1974 to 5 October 1975.

o o o *

GqbDIer
Knob

• Capture site
*■ Spring relocations I3t)
° Summer relocations C 8 )

• ''•'I I relocations ( 6 )

• Winter relocations (36)

Figure 19. Locations of elk GKl fr»)m 4 January 1974 to 4 October 1975.

National Park Service

park, was relocated 81 times (Fig. 19). She moved to
an area just south of Beaver Creek on 9 January and
was seen associated with elk marked in the Beaver
Creek region. She again was relocated in association
with Beaver Creek elk north of Beaver Creek on 10
and 1 1 January. This elk was not relocated again until
16 January, when she was in the Gobbler Knob region.
She remained in the Gobbler Knob region until the
morning of 6 February, when she was seen just south
of Beaver Creek with GK2 and BC2. On the evening
of 6 February, and on both the morning and evening
of 7 February, she remained close to Beaver Creek
and was seen associated with elk GK2, BCl, and BC2.
On 8 February, she left the Beaver Creek elk and was
relocated in a wooded area about a mile northwest of
park headquarters. She returned to the southwestern
corner of the park on 13 February and remained in
this area until 4 April (except for a brief period spent
near Gobbler Knob in early March). On this date, this
elk and elk GK2 were seen as part of a group of 12
just west of the park boundary fence in the Black Hills
National Forest. Elk GKl remained in the national
forest until 2 October 1974, when she returned to the
park (except for 2 days in later April when she was
seen inside the park). Elk GKl and GK2 were seen
with groups of elk ranging in size from 5 to 16 animals

while in the national forest. Elk GKl remained in the
Gobbler Knob region throughout the winter of 1 974-75,
except for one relocation made in the Beaver Creek
region just south of Beaver Creek near the park's west-
ern boundary on 9 December 1974. She was not as-
sociated with Beaver Creek elk at this time. This elk
was not seen during the summer of 1975 but was last
seen on 4 October 1975 in the Gobbler Knob region.
Her radio stopped functioning sometime after 29 July
1974.

Elk GK2, the adult cow marked on 18 January 1974
near Wind Cave Canyon, was relocated 54 times (Fig.
20). On 20 January 1974, just 2 days after being marked,
she was seen with elk BCl and BC2 about 1.25 miles
southeast of the fire tower, in the Beaver Creek region.
She remained in the Beaver Creek region, and asso-
ciated with Beaver Creek elk, until sometime between
7 and 12 February 1974. On one occasion she was
relocated as far north as 0.4 km (0.25 mile) north of
the fire tower. On 13 February, she was seen again in
the Gobbler Knob region, and remained there until 4
April 1974, when she was seen with elk GKl and 10
others, outside the park in the Black Hills National
Forest. Elk GK2 was relocated only in the national
forest until 19 August 1974, the date she was last
located.

oo o

Gc bDler
A Knob

* Capture ".ite

* Spr ing relocal ions (21)
o Summer f eloca I ions (11 )

• Fa 11 relocations ( )

• Winter relocations (22)

Figure 20. Locations of elk GK2 from 18 January to 19 August 1974.

Natural Resources Report No. 13

Group Constancy

The degree of association between any two individ-
uals was expressed by using the following formula for
the coefficient of association:

2cib
A+B
where A is the number of times elk A was observed
throughout a particular period of time, B is the number
of times elk B was observed throughout the period,
and ah is the number of times that A and B were seen
together throughout the period. Theoretically, the
coefficient of association would be if two individuals
are never seen together and 1 if they are always ob-
served together. Since elk were marked at different
times throughout the study, coefficients of association
for any set of two individuals were calculated where
both elk were known to be marked for the same period
of time. Thus, the periods of time used for which coef-
ficients of association were computed varied between
sets of individuals. Knight (1970), who used this
method to determine elk group constancy in the Sun
River herd of Montana, found that mean coefficients
of association were relatively low. He reported that
groups of elk seldom contained the same individuals
all the time. He interpreted a coefficient of association
lower than 0.5 as meaning that the attraction or as-
sociation between particular elk within the population
was not great.

In this study, coefficients of association between
individuals within a region were relatively high com-
pared with the degree of association between individ-
uals of different regions (Table 3). The degree of as-
sociation for elk marked in the Boland Ridge region
varied from 0.42 to 0.68 (x = 0.53). The highest degree
of association occurred between elk BRl and BR3.
Both of these elk had a somewhat lower association
index with elk BR2. Elk from this region were never
seen associated with elk marked in any other regions
of the park.

Table 3. Coefficients of association for marked members of
cow-calf herds in Wind Cave National Park.

GKl GK2 BCl BC2 BC3 BC4 BC5 BRl BR2 BR3

GKl
GK2
BCl
BC2
BC3
BC4
BC5
BRl
BR2

0.63 0.04 0.05

0.12 0.\5

0.44 0.43 0.24 0.29

0.38 0.38 0.33

0.12 0,54

0.08

0.42

0.68
0.49

Elk marked in the Beaver Creek region had coeffi-
cients of association ranging from 0.08 to 0.54 (x =
0.32). Except for BC3 and BC5, there seemed to be
no great attraction between marked individuals in this
region. However, marked elk were never seen outside
the region, and individuals commonly were seen as-
sociated in different combinations. Elk of this region,
as well as other regions, occurred in several groups
that changed in composition as they intermingled and
individuals left one group and joined another for vary-
ing periods. Associations between Beaver Creek and
Gobbler Knob elk were relatively low, ranging from
0.04 to 0.15 (x = 0.09) for those pairs where associa-
tions occurred. Coefficients of association may have
been even lower had we not observed intensively the
movements of the Gobbler Knob elk while they were
in the Beaver Creek region.

The coefficient of association for Gobbler Knob elk
GKl and GK2 was relatively high (0.63).

Bull groups changed in composition often, and no
lasting associations were detected between particular
marked individuals. Coefficients of association for
marked bulls were relatively low, ranging from 0.04 to
0.28 (x = 0.12), for those pairs where associations
occurred.

AREAS OF INTENSIVE USE

The areas used most intensively by all elk in any
particular region of the park were closely similar to the
areas used by individuals marked in that region.
Cow-calf groups in the Boland Ridge region used areas
on or east of Boland Ridge intensively during all sea-
sons, and the area west of Red Valley adjacent to
Highland Creek most intensively during summer and
less intensively during other seasons (Fig. 21 and 22).

Ranges used by cow-calf groups in the Beaver Creek
region were similar for all seasons, with somewhat
low er usage of the forested area betw ecn Beaver Creek
and Sanctuary prairie dog town during fall and winter
and the area south of Beaver Creek during spring and
summer.

Ranges used by cow-calf groups of the Gobbler
Knob region were similar during spring and fall.

National Park Service

' ' 1 1

1 mile
\ Primary Range

Secondary Range

Elk not seen

Figure 21. Primary and secondary ranges used by cow-calf
groups during spring 1974 (top), and summer
1973 and 1974 [ho itom).

Groups in this region used the park's extreme south-
western corner most intensively, although use of this
region was less during summer, probably because elk
moved to areas west of the park (Figs. 19 and 20).
Areas in or near Bison Flats and Gobbler Knob were
used more intensively during winter than during the
rest of the year.

Bulls generally used the northern half of the park
much more intensively than the southern half during
all seasons (Figs. 23 and 24). and made little use of the
park's southwestern portion during winter and spring.
The area east of Red Valley was used less intensively
during spring than during the rest of the year. Ranges
west of Rankin Ridge were used less during spring and
winter than during summer and fall. Bulls used areas
in and near Sanctuary prairie dog town (Fig. 3) and
just south of Beaver Creek most intensively during fall,
and less intensively during spring.

Secondary Range

Elk not seen

Figure 22. Primary and secondary ranges used by cow-calf
groups during fall 1973 and 1974 (/(^p), and winter
1973-74 and 1974-75 (bottom).

Factors Influencing Areas Used

Elk were seen most often during several hours after
sunrise and before sunset (Fig. 25). Few elk were seen
during midday.

Craighead et al. (1973) found that most of an elk's
daily activities in Yellowstone National Park comprise
either feeding (44%) or bedding (46%). During this
study, most elk were seen feeding at sunrise, and few
were bedded (Fig. 26). As the morning progressed,
feeding declined and bedding increased. The percent-
age of elk seen feeding generally increased again in
midafternoon and reached a second daily peak at sun-
set. Elk used grassland most intensively during hours
closest to sunrise and sunset, and least at midday.
Forest areas were used more during hours closest to
midday (Fig. 27). Feeding activity generally occurred
in grassland areas; 91% of all feeding elk were in grass-
land and 9% were in forest. In some studies, elk have

Natural Resources Report No. 13

Primafy Range
Secondary Range
E I k not seen

Primary Range

;;;| Secondary Range

Elk not seen

Figure 23. Primary and secondary ranges used by adult
bulls during spring \974(iop). and summer 1973
and 1974 (bottom).

- 30-

o
I

- 20H

-U 10-

— I 1 1 1 1 r-

s y< 1 2 } 4+ -4

"1 1 1 1 —

3 ? 1 SS

Figure 25. Numbers of elk seen per hour for each of the 4
hours after sunrise (SR) and before sunset (SS) at
Wind Cave National Park.

[VX] Primary Range
• j • I Secondary Range

Elk not seen

Figure 24. Primary and secondary ranges used by adult
bulls during fall 1973 and 1974 (top), and winter
1973-74 and 1974-75 (hot torn).

lOOn
90
80
70'

|6o^

'^ 50

o
<^ 40

'OH

— 1 1 1 1 1

SH I 2 3 4 +

"T ' ' ' — rr-r"

-4 3 2 1 b^

IIdli I

Figure 26. Percentages of all elk seen feeding or bedded
during early morning and late afternoon hours at
Wind Cave National Park.

National Park Service

100
90
80
70H

^ 60

a> 50-
o

I 40-
30-
20

lOH

Table 4. Percentages of all elk seen in either forest or grassland
during the four seasons at Wind Cave National Park.

-4

3 2 1 SS

Hour-

Figure 27. Percentages of ail elk seen in either forest or
grassland during early morning and late after-
noon hours at Wind Cave National Park.

been found to prefer grasses ( Baldwin and Patton 1 938;
DeNio 1938; Gaffney 1941; Harper 1962; Morris and
Schwartz 1957). In contrast to feeding elk, 60% of the
bedded elk were seen in forest and 40% in grassland.
Elk of the Boland Ridge region seemed to bed in grass-
land areas more than did elk of the other two regions,
perhaps because these elk usually were not near public
roads and thus were seldom disturbed by park visitors.
More elk were seen in grassland areas than in forest
(Table 4), probably because elk were seen more easily
in grassland and also because grassland constituted a
larger proportion of the park, and thus more time was
spent observing in this type of habitat. Use of grassland
areas differed only slightly during the four seasons, but
use of grassland was above average during spring and
fall. Grassland areas where prairie dog towns occurred
generally received highest use during fall and little use
during winter. Percentages of all elk (2,395) seen on
prairie dog towns during the four seasons were as fol-
lows: spring, 20.2; summer, 32. 1 ; fall, 43.0; and winter,
4.7. Whether this was because elk used prairie dog
towns or because prairie dogs occupied prime, level,
grassland areas, often near forest, is unknown. Elk
used prairie dog towns extensively as rutting areas,
which explains some of the heavy fall use.

Season

No.

Forest

Grassland

Spring

4,235

8.9

91.1

Summer

5,477

15.4

84.6

Fall

6,171

111

88,9

Winter

8,029

16.1

83.9

Total or ave

rages

23,912

12.9

87.1

Elk preferred to use areas of level ground (Tables
5 and 6) and gentle slopes (0-25°) during all seasons,
but used steeper slopes more during fall and winter
than during other seasons (Table 5). The preferences
of elk for different exposures were ranked as follows
(from highest to lowest): east, south, west, and north
(Table 6). Winter was the only season in which use of
all exposures was nearly equal. Use of eastern expo-
sures declined in summer. Elk used southern exposures
uniformly during all seasons. Western exposures were
used relatively more during fall and winter and northern
exposures during winter. Snow depth was seldom, if
ever, great enough to influence feeding. Mackie (1970)
found that (1) use of northerly exposures by elk in the
breaks of the Missouri River, Montana, was especially
intensive during summer and winter; (2) western ex-
posures received greater use in winter; (3) cold wind
and deep snow, that prevailed in winter, did not seem
to inhibit the use of northerly exposures; (4) slopes
facing south, southeast, and east received relatively
high use during spring; and (5) southern and south-
western exposures were used intensively in fall.

Table 5. Percentages of all elk seen on slopes of varying degrees
of steepness during the four seasons at Wind Cave Na-
tional Park.

Degrees of slope

Season

No.

0-25°

>25°

Spring

4,208

72.4

27.6

Summer

5,462

73.3

26.7

Fall

6,112

69.0

31.0

Winter

7.934

63.8

36.2

Total or averages

23,716

69.6

30.4

Natural Resources Report No. 13

Table 6. Percentages of all elk seen on level ground or slopes with various exposures
during the four seasons at Wind Cave National Park.

Season

Exposure

No.

Level

Spring 4.224 38.3 6.5 18.7 23.4 13.1

Summer 5,470 39.7 11.9 19.2 15.6 13.6

Fall 6.125 25.3 9.4 19.1 27.4 18.8

Winter 8.439 25.0 19.0 19.1 19.3 17.6

Total or averages 24.258 32.1 11.7 19.0 21.4 15.8

DISCUSSION AND
MANAGEMENT IMPLICATIONS

Wind Cave National Park was occupied by three
relatively discrete cow-calf herds and, in all probabil-
ity, also by relatively discrete bull herds. Other re-
searchers have reported the presence of multiple herds
in small, specified areas but none as small as WCNP
(114 km^ or 44 square miles). McCullough (1971) found
that Tule elk in Owens Valley, California (1,942 km'
or 750 square miles) occurred in five relatively discrete
herds, and Troyer (1960) noted that Roosevelt elk in-
troduced to Afognak Island, Alaska (2,020 km^ or 780
square miles) were also divided into five major herds.
Both researchers found little interchange of individuals
between herds.

Group composition at WCNP changed through time.
Others have found that group composition is dynamic
and that elk do not necessarily remain in well-defined
groups indefinitely (Craighead et al. 1973; Harper 1964,
1966; Knight 1970; Mackie 1970; Moran 1973; Struhs-
aker 1967). Franklin et al. (1975) found composition
of a group of non migratory Roosevelt elk to be con-
stant, even when the group temporarily became as-
sociated with other groups. They speculated that this
behavior is related to group development; large groups
evolve stronger social ties and are more stable than
smaller, less strongly developed groups.

Researchers studying elk migration patterns in or
near Yellowstone National Park have shown that elk
herds generally occupied traditional winter and sum-
mer ranges and that only a few individuals moved be-
tween herds (Brazda 1953; Cole 1969; Craighead et al.
1972; Houston 1974; Murie 1951). Schwartz and Mitch-
ell (1945) reported that both resident and migratory
Roosevelt elk in Washington used traditional wintering
areas, and Knight (1970) reported use of traditional

winter ranges by elk from the Sun River herd in
Montana.

Areas used by cows and calves of the Beaver Creek
region have shifted during recent years. Before heli-
copter trapping operations were begun in 1970, cows
and calves frequently were seen in the vicinity of the
corral. During 1969 and 1970, 80 elk inadvertently
wandered into the trap and were captured. During the
present study, however, cow-calf groups were seldom
seen within a mile of the structure. This change in
movements may have been a natural shift because the
group of cows and calves that occupied the area near
the trap were captured and shipped out, or the cows
may have learned to fear the trap.

Before this study, WCNP's elk population was cen-
sused periodically and reduced to prevent overuse of
the range. Elk from the Boland Ridge region were
caught most easily during helicopter drives because
trap wings open to the east, toward that region (Fig.
1). Elk west and south of the trap were more difficult
to capture because they had to be driven long distances
through rough, forested terrain, around the wings, and
back west into the trap. Nevertheless, many elk from
the Beaver Creek region were trapped (Lovaas 1973a).
Elk have never been driven successfully to the trap
from the Gobbler Knob region.

The total elk population in March 1 975 w as estimated
to be 450-500 animals (Lovaas 1975). based on both
aerial and ground counts. Although no census was
made in the winter of 1975-76, elk numbers during this
time were reportedly higher than during the previous
winter, and the Beaver Creek cow-calf herd exceeded
200 animals. To attain the population size recom-
mended in the park's wildlife management plan

National Park Service

(150-300), the population had to be reduced.

Information from this study, indicating the presence
of three separate cow-calf herds in the park, was used
in elk trapping operations during January 1976 and
1977. In 1976, a herd reduction of up to 210 cows and
calves was planned, with the helicopter costs to be
paid by several Indian tribes who were to receive the
elk. In a memorandum of understanding between the
National Park Service and the tribes, a quota of 50
cows and calves was established for the Boland Ridge
herd and 160 for the Beaver Creek herd. If more than
the quota were captured from either herd, excess an-
imals were to be released to prevent overreduction of
that herd. Ail bulls captured also would be removed
from the park and would not count against the quotas.
Twenty-two elk were captured from the Boland Ridge
region and 108 from the Beaver Creek region. Two elk
died during the trapping operation, and the remaining
128 were divided among four Indian tribes and were
shipped to their reservations by truck.

In 1977, the same agreement was used with five In-
dian tribes, except that the quotas were 70 cows and
calves from the Boland Ridge herd and 230 from the
Beaver Creek herd. One hundred and fifty-four cows
and calves were trapped from the Beaver Creek region
and 41 from the Boland Ridge region. Only 185 were
shipped; 3 died, I reacted positively to brucellosis, and
6 were released.

Little effort was spent in trying to trap adult bulls
because they had proved to be difficult to capture in
previous helicopter drives (Lovaas 1973b). Those oc-
curring with cow-calf groups were often captured.

Elk numbers in the Gobbler Knob region declined
in recent years. The total herd count was 74 in 1970
(Lovaas 1970), but only 46 in 1974 and 41 in 1975.
Hunting pressure in the area just west of the park
evidently increased, as determined by an examination
of hunter report cards sent to the South Dakota De-
partment of Game, Fish, and Parks from 1971 to 1974.
Several elk were killed within 6.4 km (4 miles) of the
park's western boundary during the September hunting
seasons of 1971 through 1974 (Table 7). Because at
least a portion of the Gobbler Knob herd remains in
the area during the hunting season (e.g., elk GKl),
some of the Gobbler Knob elk were probably taken
by hunters. Poaching was also a factor. Because this
herd was declining, no reductions were contemplated
for that region. Periodic censusing will be used to de-
termine if future reduction will be required.

According to National Park Service policy (U.S.
Department of the Interior 1975), regulation of wildlife
populations by natural means is preferred. However,
when natural means are not feasible, live trapping is
the preferred method of direct management, and public
hunting outside the boundary of the park is encouraged.

Natural control of elk numbers in WCNP is not
feasible because numbers and kinds of predators found
in this relatively small, fenced park are insufficient to
regulate the elk population. However, movements of
elk into the Black Hills National Forest, a public hunt-
ing area adjacent to the Gobbler Knob region, is en-
couraged by the presence of the 122-cm (4-ft) fence in
the park's southwest corner. Low fences along other
boundaries of the park are not practical because elk
could invade private lands. According to National Park
Service policy, therefore, the elk population at WCNP
is being regulated by those means considered most
desirable — hunting outside the boundary and trapping
and transplanting. If an additional trap is needed for
future management of the Gobbler Knob and Beaver
Creek elk, it could be built between the Beaver Creek
and Gobbler Knob regions. Wind Cave Canyon (Fig.
I) would provide a site hidden from public view but
near a road that would allow easy removal of elk by
truck. It is close to areas used by elk in winter. Other
possible trap sites are east of Lookout Point on the
south side of Beaver Creek and the grassland area in
and adjacent to Sanctuary prairie dog town. This last-
named location may be undesirable, however, because
of its close proximity to woods, which may be detri-
mental to effective elk trapping, and its poor location
in relation to Gobbler Knob elk. Experimental drives
should be conducted before trap construction to make
sure that elk can be driven to any new trap site.

Table 7. Number of elk reported killed during the September hunt-
ing seasons of 1971 through 1974 within 6.4 km (4 miles)
of the western boundary of Wind Cave National Park and
for the remainder of the southern Black Hills region.

Additional kills reported

for the entire southern

Black Hills region

34
53
32
33

Bulls could be reduced by baiting them into a trap, as
we did during this study. Additional traps would be
needed to capture bulls from the Boland Ridge and
Gobbler Knob regions. Daike etal. (1965) showed that
elk use salt most readily during May and June. We
found that bulls also were captured during late fall and
winter by using salt as bait.

Range studies are needed in each of the three regions
of the park to determine more adequately optimum
herd size for each region. A better understanding of
the relationship of elk to other herbivores in the park
also is needed.

No. of re

;ported kill

near

the park

Year

Bull:

Cows

Calves

Total

1971

1972

1973

1974

Natural Resources Report No. 13

REFERENCES

Baldwin, W. P., and C. P. Patton. 1938. A preliminary
study of the food habits of elk in Virginia. Trans. N. Am.
WilJI. Conf. 3:747-755.

Beetle. A. A. 1970. Recommended plant names. Univ.
Wyo. Agric. Exp. Stn. Res. J. 31. 124 p.

Brazda, a. R. 1953. Elk migration patterns and some of the
factors affecting movements in the Gallatin River Drain-
age, Montana, y. WiUll. Manage. 17(l):9-23.

Cole. G. F. 1969. The elk of Grand Teton and southern
Yellowstone National Parks. Research Rep. GRTE-N-I.
Yellowstone National Park. 192 p.

Craighead. J. J.. M. G. Hornocker, M. W. Shoesmith,
and R. I. Ellis. 1969. A marking technique for elk. J.
Wildl. Manage. 33(4):906-909.

Craighead, J. J., G. Atwell, and B. W. O'Gara. 1972.
Elk migrations in and near Yellowstone National Park.
Wildl. Monogr. 29. 48 p.

Craighead, J. J., F. C. Craighead, Jr., R. L. Ruff, and
B. W. O'Gara. 1973. Home ranges and activity patterns
of nonmigratory elk of the Madison Drainage herd as de-
termined by biotelemetry. Wildl. Monoi^r. 33. 50 p.

Dalke, p. D., R. p. Beeman, F. J. Kindel, R. J. Robel,
and T. R. Williams. 1965. Use of salt by elk in Idaho.
J. Wildl. Manaiic. 29(2):3 19-332.

DeNio, R. M. 1938. Elk and deer foods and feeding habits.
Trans. N. Am. Wildl. Conf. 3:421-427.

DusEK, G. L. 1975. Range relations of mule deer and cattle
in prairie habitat. J. Wildl. Manatee. 39(3):605-6l6.

Franklin, W. L., A. S. Mossman, and M. Dole. 1975.
Social organization and home range of Roosevelt elk. ./.
Mammal. 56(1); 102-1 18.

Gaffney, W. S. 1941. The effects of winter elk browsing,
south fork of the Rathead River, Montana. J. Wildl. Man-
ai^e. 5(4):427-453.

Harper, J. A. 1962. Daytime feeding habits of Roosevelt
elk on Boyes Prairie, California. ./. Wildl. Manaf^e.
26(1): 97- 100.

. 1964. Movement and associated behavior of Roose-
velt elk in southwestern Oregon. Proc. Anna. Conf. West-
ern Assoc. State Game and Fish Comm., San Francisco,
Calif. 44:139-141.

. 1966. Ecological study of Roosevelt elk. Game Res.

Rep. No. 1. Res. Div., Oregon State Game Comm. 29 p.

Heezen, K. L., and J. R. Tester. 1967. Evaluation of radio-
tracking by triangulation with special reference to deer
movements, y. Wildl. Manai>e. 31( 1): 124-141.

Hipschman, D. 1959. A look at the past half century. S. D.
Conserv. Dii>est 26(2):20-31, 37.

Houston, D. B. 1974. The northern Yellowstone elk, parts
I and II — history and demography. Yellowstone National
Park. 185 p.

Johnson, D. E. 1951. Biology of the elk calf, Cerviis can-
adensis nelsoni. .1 . Wildl. Manai^e. 15(4):396-410.

Knight, R. R. 1966. Effectivenessof neckbands for marking
elk.i. Wildl. Manage. 30(4): 845-846.

. 1970. The Sun River elk herd. Wildl. Monogr. 23.

66 p.

Lovaas, a. L. 1970. Typewritten memorandum of 13 March
1970 to Superintendent of Wind Cave National Park, re-
porting on aerial elk census.

. 1973a. A cooperative elk trapping program in Wind

Cave National Park. Wildl. Soc. Bull. 1(2):93-100.

— . 1973b. Prairie dogs and black-footed ferrets in the

national parks. Pages 139-148//; R. L. Linder and C. N.
Hillman. eds. Black-Footed Ferret and Prairie Dog Work-
shop. South Dakota State Univ., Brookings.
. 1975. Typewritten memorandum of 19 March 1975

to Superintendent of Wind Cave National Park, reporting
on aerial elk census.

Lovaas, A. L., and P. T. Bromley. 1972. Preliminary stud-
ies of pronghorn antelope — blacktail prairie dog relations
in Wind Cave National Park. Pages 1 15-156 /« Proc. 5th
Bienn. Antelope States Workshop, Billings, Mont.

Lovaas, A. L., J. L. Egan, and R. R. Knight. 1966. Aerial
counting of two Montana elk herds. J. Wildl. Manage.
30(2):364-369.

Mackie, R. J. 1970. Range ecology and relations of mule
deer, elk. and cattle in the Missouri River Breaks, Mon-
tana. Wildl. Monogr. 20. 79 p.

McCuLLOUGH, D. R. 1971. The Tule Elk, Its History, Be-
havior, and Ecology. Univ. of Calif. Press. Berkeley. 209

P-
Moran. R. J. 1973. The Rocky Mountain Elk in Michigan.

Michigan Dept. of Nat. Resour. R and D Dep. No. 267.

93 p.
Morris, M. S., and J. E. Schwartz. 1957. Mule deer and

elk food habits on the National Bison Range. J. Wildl.

Manage. 21(2): 189^193.
MuRiE, O. J. 1951. The Elkof North America. The Stackpole

Co., Harrisburg, Pa. 376 p.
Petersburg. S. J. 1973. Bull bison behavior at Wind Cave

National Park. M.S. Thesis. Iowa State Univ., Ames. 302

P-

Progulske, D. R. 1957. A collar for identification of big
game.y. Wildl. Manage. 21(2):25I-252.

Russo, J. P. 1964. The Kaibab North Deer Herd, Its His-
tory, Problems and Management. Ariz. Game Fish Dep.
Wildl. Bull. 7. 195 p.

Schwartz. J. E.. and G. F. Mitchell. 1945. The Roosevelt
elk on the Olympic Peninsula. Washington. 7. Wildl. Man-
age. 9(4):295-3l9.

Shult. M. J. 1972. American bison behavior patterns at
Wind Cave National Park. Ph.D. Thesis. Iowa State
Univ., Ames. 178 p.

Smuts. G. L. 1974. Game movements in the Kruger National
Park and their relationship to the segregation of sub-pop-
ulations and the allocation of culling compartments. J. S.
Afr. Wildl. Manage. Assoc. 4(l):51-58.

Struhsaker, T. T. 1967. Behaviour of elk (Cervns cana-
densis) during the rut. Z. Tierpsychol. 24:80-1 14.

Natural Resources Report No. 13

TROYtR, W. A. I960. The Roosevelt elk on Afognak Island. Vari and. K. L. 1976a. Herd organization and movements

Alaska, y. Wildl. Manage. 24(1):I5-21. of elk in Wind Cave National Park. South Dakota. M.S.

U.S. Departmknt of the Interior. 1975. Management Thesis. Iowa State Univ., Ames. 153 p.

Policies, National Park Service. 125 p. . 1976b. Techniques for elk immobilization with suc-

U. S. Soil Conservation Service. 1969. Conservation cinylcholine chloride. Proc. Iowa Acad. Sci. 82(3-4)

plan for Wind Cave National Park. 46 p. 194-197.

ACKNOWLEDGMENTS

The cooperation of many people made this project possible, and we wish to express our
appreciation to the following:

Drs. Arnold O. Haugen, Leader, and Kenneth R. Russell. Assistant Leader. Iowa
Cooperative Wildlife Research Unit, who helped to initiate the project and provided ex-
pertise at the beginning stages of work; the staff at Wind Cave National Park (special
thanks to Lester McClanahan, Superintendent, and seasonal rangers Thomas Bean, Doug-
las Buehler, and Larry Frederick, for their assistance); Jerry D. Andrew. Iowa State
University, for field and clerical assistance during the latter portion of the project; David
Schlapki)hl, Nancy Smith, Lorraine Martzahn, and Robert Major, Iowa State University,
for their clerical help; John May and Drs. William K. Seitz. Michael K. Petersen, and
David Graham. Iowa State University, who helped immobilize elk for marking; Dr. Richard
Knight. National Park Service, for expert consultation before we began using a helicopter
to pursue elk for immobilization and marking; Dr. David F. Cox. Iowa State University,
for assisting with computer programming; Dr. Frwin E. Klaas. Iowa Cooperative Wildlife
Research Unit, who provided helpful editorial comments; and Hazel Clausen. Iowa Co-
operative Wildlife Research Unit, for typing the manuscript; personnel of the South Dakota
Department of Game. Fish and Parks for providing ear tags and hunter report cards; and
Carolyn Varland for her field and clerical assistance.

Ihis project was funded by the National Park Service and the U.S. Fish and Wildlife
Service and was administered by the Iowa Cooperative Wildlife Research Unit, supported
jointly by the U.S. Fish and Wildlife Service. Iowa State University of Science and
lechnology, Iowa State Conservation Commission, and the Wildlife Management Instiliilc.
Journal Paper No. J-8912 of the Iowa Agriculture and Home Economics Experiment
Station; Project No. 1998.

* U. S. GOVERNMENT PRINTING OFFICE : 1978 — 722-802

National Park Service

DATE QUE 1